Computerized subtraction radiography is a potentially useful tool for assessing change in alveolar crestal bone status with greater accuracy than traditional direct comparison radiographs. Its limitation is that it requires reproducible geometry for radiograph series and if this requirement is not met the resulting subtraction image may be uninterpretable. An objective measure of interpretability will serve to enhance the confidence a researcher or clinician has in interpreting a subtraction image. It will also serve as means for laboratories to compare results. Because the texture patterns of subtraction images are influenced by the angle difference between radiographs, texture analysis of subtraction image may provide an objective measure of its interpretability. This study will evaluate texture analysis of a subtraction image as a predictor of the probability that a crestal site will be correctly interpreted through subtraction radiography. It will also evaluate the influence that radiographic quality may have on texture analysis as a predictor of correct interpretation.